Cao, Y., D.P. Larsen, and R.St-J. Thorne. 2001. Rare species in multivariate analysis for bioassessment: some considerations. J. N. Am. Benthol. Soc. 20(1)144-153.
Bioassessment is increasingly contributing to the recovery, protection and management of freshwater resources. Field survey data typically contains a large number of rare species -- simply because the majority of species in natural communities are rare. How we handle these rare species is often confusing and controversial, but how we handle rare species may be directly related to our ability to detect and reliably measure human impacts. This paper begins by reviewing the definition of rarity and its relationships with sample size, local species abundance and habitat characteristics. Then, it moves to a discussion of the role of rare species in multivariate analysis and bioassessment. We conclude that 1) the importance of rare species is related to the objectives and spatial scale of the study, as well as to the definition and severity of impact; 2) abundant species are often associated with major environmental gradients and these gradients are often detected with multivariate techniques, however the theoretical or empirical justification for excluding rare species in bioassessment is weak. We recommend that sample size and the rules for excluding rare species before conducting multivariate analyses need to be evaluated carefully for their unintended influences on the outcome of the analysis. Improper choice may compromise the sensitivity and reliability of the multivariate analysis in bioassessment.